Despite the impressive accumulation of information on the structure and function of transfer RNAs in recent years, relatively little is known about their biosynthesis. This situation stems in large part from the inherent instability of the biosynthetic intermediates in the bacterial and mammalian systems investigated to date. A more promising approach would invoke implementation of a system in which the biosynthetic intermediates are normally detectable and there is the opportunity to perform a genetic dissection of the various biosynthetic events. The bacteriophage T4 induces the synthesis of eight transfer RNAs upon infection of Escherichia coli. A unique feature of this system is the existence of a number of transfer RNA precursor molecules, which can be readily obtained in reasonable amounts and purity. Three different precursor transfer RNA molecules have been identified to date, each containing the sequences of two mature transfer RNA species as well as additional nucleotides at the 5' and 3' termini and in the intertransfer RNA spacer region. We propose to determine the complete nucleotide sequences of two of these transfer RNA precursor molecules. This information, coupled with an analysis of available biosynthetic mutants of both host and virus, should allow the identification of the specific features of the precursor molecules that insure their accurate biosynthesis.